Korg Forums

Ill try and make this make sense because its a little hard to explain what I'm trying to maybe do, and maybe it wont really work as expected:

Has anyone ever tried putting a quality stereo speaker close to the R3 microphone, and tried playing audio of a 'weird sound' could be wavetable, granular, just a single note with lots of modulations happening [from another sound source, i have a few vsts - even just random audio]- coming through the speaker, right infront of the R3 mic, and you have it set to record formant motion - while playing a note on the R3- to kind of get vocoding of whats coming through the speaker -

question is, if you do that, would it modulate in a cool weird way, or just sound like the formant shape made from the speaker, en lou of mouth shape you would make anyway, like, basically the same as 'hello', from the lets say edgy serum bass wavetable, would you hear that too or just the R3 oscillators + formants, and basically just make the internal R3 oscillators modulate with the formant shape of whatever is going into the mic.

Sound wise, I have the R3 going into my mackies, but i also have a tiny table set of 2 blackstar amps, behind the R3. I dont know if I need both to be on the same speakers for this to 'work' haha. I am definitely gonna try it, just out of sheer curiosity, but my speaker cant really move far enough from its current setup, so ill have to carry the r3 over to the speaker for this experiment. Not sure soundwise it would produce anything like I'm hoping, in trying to get a really outlandish sound that way, but it would take a little time re-arranging my studio setup just to snake the wires close enough to the speaker. Its only 2 feet away! oh well. experiment time!

I already get a lot of awesome sounds this way via the rear inputs, just trying to 'push' it a bit more. And recently got a nice setup with a nanokey studio, although i still have to program the knobs with midi CC. [separate issue]

Hi,

You can record up to 7.5s formant data.

If you use a simple vocoder program and your speaker + R3 mic setup, the sound that you play on the speaker will modulate the notes that you play on the R3.
If you record the sound that you play as formant motion, you can use that recording to modulate the notes that you play over and over again. But it will sound the same as if you would play it in real life again through the mic.

The analyzer filters of a vocoder are basically a low rate spectrum analyzer (roughly every 20ms, an output for 16 fequency bands). That spectrum is then applied to the notes (kind a like an equalizer) (= the modulation).
The formant motion data that is recorded is the output of the analyzer filters.

So snippets like 'hello' will keep sounding as 'hello'. But more dynamic and spectrum rich snippets will be heavily reduced and will loose detail (50Hz data, only on 16 frequency bands). Though the 'envelope' of your original sound should be recognizable.

Also, why the setup of speaker and mic? I think you can just run audiocable straight into the R3.

Have fun with experimenting time.

Hello there, and yes, I actually have two other desktop synths going into each of the audio inputs.

I guess i'm attempting some sort of lo fi effect applied to unusual incoming sound from the speaker, lets say a babbling brook or maybe a train whistle, or perhaps other 'ordinary' sounds, close enough to the mic like a voice might do - and perhaps apply waveshaping, such as bit crushing or the pickup or some other. Not to granulate it per se, but i guess, to 'vocode' something other than my voice, kind of -

I guess that was why, really. Mostly for pushing unusual sound experimentation. whether it yields anything Ill find out soon , I guess! Thanks for the info - so far with regular vocal recording, i get this weird pause at the end, but ive only really just begun to use the vocoder. I can do the basic simple stuff, which sounds great, really. And comes through nice and clear.

Update on my vocoder / serum experiment. I would say its been successful, but not in the way originally anticipated. Definitely I can say some aspect from the recorded wavetable from serum that I had made and was collecting some 'dust' for a bit, paid off in terms of what I was wanting from it, but only when I changed the formant shift and offset, and base osc waveform on the R3. I ended up using my sound link mini to the vocoder mic.

I also realized, later on - that I can even apply waveshaping and lfos, mod matrix assignments, motion sequencing...effects... basically like any other sound - AND add the vocoder formant ON TOP of that.. . suffice to say I'm pretty stoked on how it turned out. Until now I thought that was impossible.

Ive only recorded one waveform as a test., the only caveat , as has been mentioned previously, the resolution is not near the original, but, if i shift the vocoder parameters around a bit, the original wavebank can actually be 'almost' heard well. What the good news is though, besides the above, is the resulting waveform sounds like a new kind of sound. That sounds wavetable-y and FMy at the same time. But definitely good for bass stuff, for sure.

I selected osc1 as pwm, with the vpm mode, and used the harmonic parameter to apply motion sequencing, added in osc 2 as ring mod type, didnt apply every conceivable mod slot [yet], since it could muddy the wave too much.

But I plan to find tune it to where I can make 7 more of these [i like the inbuilt vocoder waveforms but I think i accidentally overwrote one, so I might have to get into the editor and try getting the factory sounds and get that back in again. But im pleasantly surprised.

Good!
Thanks for the update.
Have more fun!